Currently, many companies that generate a large amount of outgoing mail have their mail verified and accepted by the United States Postal Service (“USPS”) at their mail production facilities rather than delivering it to a post office for acceptance processing. These in-house verification and acceptance centers are often referred to as Detached Mail Units (“DMU”). Typically, the company will present one or more mailings to a DMU where they will be processed and sent for delivery. These companies are responsible for reporting information regarding the mailing to USPS and for insuring that the mailing conforms to USPS standards. This includes insuring that the barcode information, sort information, and/or rate information is accurate and conforms to USPS standards.
USPS periodically sends postal clerks to the DMUs to verify that the mailer provided accurate information and that the mailing comports with USPS standards. Typically, an automated verification device, such as a Mailing Evaluation Readability Lookup Instrument (“MERLIN”), is used to verify large amounts of mail efficiently. In most cases, rather then check the entire mailing, the clerk selects a random sample of each mailing and uses that sample to verify the entire mailing. In order for the sample to be an accurate representation of the entire mailing, however, it must be a statistically valid random sample. Providing a standard procedure for a clerk to obtain such a sample has proven difficult because of the fact that each mailing may have different properties.
For example, mailings prepared at DMUs can vary by size, location, type of mail, and ordering. With respect to size, a company may process small, medium, or large mailings. The size of the mailing alters the amount of mail need for a statistically valid random sample.
With respect to location, a mailing may reside in several locations throughout the DMU. When a clerk arrives to verify a mailing, it may be in processing, stored on the floor, or already partially delivered. These locations must also be taken into account in determining a statistically valid random sample.
With respect to type of mail, a mailer may mail letters, flats, packages, or a combination of all three. A letter mailer generally works with large volumes of letter mail. A flat mailer generally works with magazines, catalogues, and other “flat” mail. This may affect the amount of mail that needs to be sampled. Therefore, this also must be taken into account in determining a statistically valid random sample.
With respect to ordering, a mailer may also be a “list mailer.” Mail from list mailers is presorted by computer based on a predetermined order. Other mailers, however, may use optical scanners and bar code sorters. In this case, the order of the mail cannot be determined until the mail has been sorted. The ordering of the mail may also affect the choice of a statistically valid random sample.
These variations make it difficult for a clerk to apply a single random sampling method to different mailings. Therefore, it is desirable to provide a method for obtaining a random sampling that takes into account these variations.